Near-Infrared Fluorescence Probes for Enzymes Based on Binding

Apr 7, 2012 - Phone: +81-3-5841-4850. Fax: +81-3-5841-4855. ... An OFF–ON Two-Photon Fluorescent Probe for Tracking Cell Senescence in Vivo...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/ac

Near-Infrared Fluorescence Probes for Enzymes Based on Binding Affinity Modulation of Squarylium Dye Scaffold Daihi Oushiki,† Hirotatsu Kojima,‡ Yuki Takahashi,§ Toru Komatsu,† Takuya Terai,† Kenjiro Hanaoka,† Makiya Nishikawa,§ Yoshinobu Takakura,§ and Tetsuo Nagano*,†,‡ †

Graduate School of Pharmaceutical Sciences and ‡Open Innovation Center for Drug Discovery, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan § Graduate School of Pharmaceutical Sciences, Kyoto University, 46-29 Yoshida-Shimo-Adachi-cho, Sakyo-ku, Kyoto 606-8501, Japan S Supporting Information *

ABSTRACT: We present a novel design strategy for near-infrared (NIR) fluorescence probes utilizing dye−protein interaction as a trigger for fluorescence enhancement. The design principle involves modification of a polymethine dye with cleavable functional groups that reduce the dye’s proteinbinding affinity. When these functional groups are removed by specific interaction with the target enzymes, the dye’s protein affinity is restored, protein binding occurs, and the dye’s fluorescence is strongly enhanced. To validate this strategy, we first designed and synthesized an alkaline phosphatase (ALP) sensor by introducing phosphate into the squarylium dye scaffold; this sensor was able to detect ALP-labeled secondary antibodies in Western blotting analysis. Second, we synthesized a probe for βgalactosidase (widely used as a reporter of gene expression) by means of β-galactosyl substitution of the squarylium scaffold; this sensor was able to visualize β-galactosidase activity both in vitro and in vivo. Our strategy should be applicable to obtain NIR fluorescence probes for a wide range of target enzymes.

M

sensors able to visualize exotype enzyme activities involved in various biological processes. Polymethine dyes exhibit a red shift in both their absorption and emission spectra, together with a fluorescence enhancement, upon nonspecific interaction with serum proteins under biological conditions.16−18 This characteristic is undesirable when the dyes are used as labeling agents for biomolecules, for example, because it results in high background fluorescence. However, we considered that this property might be used to advantage. We hypothesized that the introduction of suitable negatively charged functional groups into the dye might suppress dye−protein interaction, consequently blocking the protein-binding-induced spectral changes. If this were the case, subsequent removal of the functional groups by specific interaction with a target enzyme would be expected to restore the dye’s protein affinity, resulting in an increase of protein binding accompanied with an enhancement of fluorescence. In this paper, we describe the development of this novel design strategy for NIR fluorescence probes and demonstrate its utility by employing squarylium, a polymethine dye that exhibits a large fluorescence enhancement upon interaction with protein, as a scaffold to obtain polymethine-based probes for visualizing alkaline phosphatase activity (ALP; widely used as a marker enzyme in Western blotting and biological assays) and for visualizing β-galactosidase activity (widely used as a

olecular imaging methods are rapidly emerging as powerful tools for biological studies, drug discovery, and clinical diagnosis.1,2 Among them, fluorescence imaging methods are generally superior in terms of sensitivity, selectivity, and ease of use, and many fluorescence probes for visualizing target biomolecules are now widely employed.3 However, for imaging in tissues or individuals in vivo, fluorochromes whose absorption and emission maxima lie in the near-infrared (NIR) region, 650−900 nm, are preferred, because this wavelength region offers minimal interference from biomolecules, low autofluorescence, and good tissue penetration, as well as low phototoxicity to cells.4,5 Polymethine dyes, such as cyanine dyes, have been employed as NIR fluorochromes for many imaging agents, for example, cyanine-biomolecule conjugates.6 However, because they are always fluorescent, low signal-to-noise ratio caused by the high background is often a problem. Therefore, fluorescence probes whose absorption and/or fluorescence spectra change upon specific reaction with biomolecules are expected to be more suitable for in vivo imaging. For example, functional cyanine probes have recently been developed to visualize metal cations, protons, endotype enzyme activities, and so on.7−14 However, the number of available molecular design methods is still limited, and new approaches are needed to develop cyaninebased NIR probes for a wider range of biological target molecules. We previously developed an oxidative stress sensor based on a novel strategy which utilizes the differential reactivity of two linked cyanine dyes.15 This time, we set out to establish an approach that would allow the development of © 2012 American Chemical Society

Received: January 11, 2012 Accepted: April 7, 2012 Published: April 7, 2012 4404

dx.doi.org/10.1021/ac300061a | Anal. Chem. 2012, 84, 4404−4410

Analytical Chemistry

Article

(final concentration of DMSO: 0.1%), and β-galactosidase was added. Fluorescence Detection on Western Blots. Materials. The kit for western-blot analysis, Multigel II mini 10 (17W), and β-ME sample treatment for Tris SDS were from Cosmobio Co., Ltd. Membrane blocking agent and ECL WB detection reagent (GE Healthcare), rat CYP3A2 SUPERSOMES (BD Biosciences), rabbit antirat CYP3A2 antibody (Nosan Corp.), goat antirabbit IgG conjugated with ALP or HRP (Cell Signaling Technology), and DDAO-Phosphate (Invitrogen Corp.) were used. Imaging systems. We used a MAESTRO in-vivo imaging system (CRI Inc., Woburn, MA) for fluorescence and an ImageQuant LAS 4000 mini (GE Healthcare) for chemiluminescence. MAESTRO was equipped with the Red filter set (excitation filter of 616 to 661 nm range and emission filter of 675 nm long-pass) and controlled with Maestro 2.10.0 software. Western Blotting. Prior to SDS-polyacrylamide gel electrophoresis (SDS-PAGE), rat CYP3A2 SUPERSOMES were diluted with β-ME sample treatment for Tris SDS, and heated at 95 °C for 5 min. Protein molecular weight standard (APRO Marker Broad range, APRO Science Co., Ltd.) (5 μL) was loaded in lanes 1 and 11, and CYP3A2 (0.2, 0.15, 0.1, 0.03, 0.02, and 0.01 pmol) was loaded in lanes 2−7 and in lanes 12− 17. After SDS-PAGE, proteins were electroblotted onto PVDF membrane using a semidry blotter. Fluorescence Detection on Blots. The blot was blocked for 60 min and then incubated with primary antibody (rabbit antirat CYP3A2 antibody diluted to 1/3000 in T-TBS) for 2 h with gentle agitation. It was washed with T-TBS (three times, 5 min each) and cut into two portions, which were further incubated separately for 2 h, one in 1/3000 T-TBS dilution of ALP-conjugated goat antirabbit IgG antibody and the other in 1/10000 T-TBS dilution of HRP-conjugated goat antirabbit IgG antibody. The blots were washed with T-TBS (three times, 5 min each), and CYP3A2 was detected with ECL western blotting detection reagent for chemiluminescence imaging. To obtain fluorescence images, 6SqmonoPhos was diluted with 2 M Tris buffer (pH 8.0) to a final concentration of 1 μM, and the blots were incubated in this substrate solution with gentle agitation for 10 min at room temperature. They were rinsed with water, and fluorescence images were acquired. Experiments Using Living Cells. Cell Culture. HEK293 cells were from RIKEN BioResource Center. To establish a cell line stably expressing β-galactosidase, pcDNA3.1/His/lacZ (Invitrogen Corp., Carlsbad, CA) was transfected into HEK293 cells with Lipofectamine 2000 (Invitrogen Corp., Carlsbad, CA). The established cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 0.1 mM MEM nonessential amino acids solution, 2 mM Lglutamine, and 10% (v/v) fetal bovine serum (all from Invitrogen Corp., Carlsbad, CA), in a humidified incubator containing 5% CO2 in air. Imaging System. The imaging system comprised an inverted microscope (IX71, Olympus) and cooled CCD camera (Cool SNAP HQ, Photometrics, Tucson, AZ), a xenon lamp (AH2-RX-T, Olympus), a 40× objective lens (UApo/340 40x/ 1.35, Olympus), and a HQ:Cy5 filter set (U−N41008, Chroma Technology Corp., Bellows Falls, VT). The system was controlled with MetaFluor 7.1 software (Molecular Devices Corp., Downingtown, PA).

reporter of gene expression). Existing design methods for obtaining dyes that show direct fluorescence enhancement after an enzyme reaction generally involve connecting a fluorochrome to a quencher via a linker, which is cleaved by the enzyme reaction, thereby releasing the quencher and activating the fluorochrome. However, this approach is not applicable to exotype enzymes, such as ALP and β-galactosidase, so our strategy of utilizing protein that is ubiquitously present in living systems to generate the fluorescence response provides a new option for designing NIR probes for such targets.

■

EXPERIMENTAL SECTION UV−Vis Spectroscopy and Fluorometric Analysis. Materials. Organic solvents for spectrometric measurements (fluorometric grade, Dojindo) were used as supplied. Fetal bovine serum (FBS) was from Invitrogen Corp., bovine serum albumin (BSA), γ-globulins from bovine blood (BGG), sodium dodecyl sulfate (SDS), and β-galactosidase (E. coli) were from Sigma-Aldrich, and alkaline phosphatase (E. coli C75) was from Takara Bio Inc.. Instruments. UV−vis spectra were obtained on a UV-1600 (Shimadzu, Japan). Fluorescence spectroscopic studies were performed on a F4500 (Hitachi, Japan). Excitation slit width, emission slit width, and photomultiplier voltage were 2.5 nm, 5.0 nm, and 700 V (400 V for BGG solution), respectively. Dyes were dissolved in DMF or DMSO to prepare stock solutions, which were diluted in aqueous buffer or FBS to a final dye concentration of 1 μM for measurements. Relative Fluorescence Quantum Yields. Relative fluorescence quantum yields (ϕFL's) were obtained by comparing the areas under the emission spectra of the test samples excited at 600 nm after subtraction of the excitation light-derived background with that of a solution of cresyl violet in methanol (ϕFL = 0.54),19 according to the following equation ϕx /ϕst = [A st /Ax ][nx 2 /nst 2][Dx /Dst ]

where st refers to the standard, x refers to the sample, A is the absorbance at the excitation wavelength (preferably less than 0.02 to ensure high precision), n is the refractive index, and D is the area under the emission spectrum on an energy scale. Measurements of Binding Parameter with BSA. Sodium phosphate buffer (0.1 M, pH 7.4) containing 13.5% w/v BSA was diluted in phosphate buffer to the required concentration. Concentrations of BSA were calculated on the basis of a molecular weight of 66 kDa. Alkaline Phosphatase Assay. Stock solution of 6SqPhos or 6SqmonoPhos was diluted to a final concentration of 1 μM in 2 M Tris buffer (pH 8.0) containing 1% w/v BSA (final concentration of DMSO: 0.1%). A suspension of 6SqPhos or 6SqmonoPhos (3 mL) was poured into a plastic cuvette, and the absorption and fluorescence spectra were measured (before). Samples were stirred gently at 37 °C, and measurement of fluorescence intensity at 660 nm (excitation: 645 nm) was started. After 1 min, alkaline phosphatase was added, and fluorescence measurement was continued for 15 min. Finally, the absorption and fluorescence spectra were measured (after). Heat inactivation of alkaline phosphatase was performed at 105 °C for 30 min. β-Galactosidase Assay. The procedure was the same as for alkaline phosphatase assay, except that 0.1 M sodium phosphate buffer (pH 7.4) containing 1% w/v BSA was used for dilution 4405

dx.doi.org/10.1021/ac300061a | Anal. Chem. 2012, 84, 4404−4410

Analytical Chemistry

Article

Imaging of β-Galactosidase Activity. HEK293 cells or HEK293 cells transfected with plasmid encoding lacZ (βgalactosidase) gene were plated in 8-well chamber slides (NALGENE, NUNC, Thermo Scientific, Rochester, NY). The culture medium was removed, and the cells were washed with Hanks’ balanced salt solution (HBSS) (Invitrogen Corp., Carlsbad, CA). 6SqGal or 6SqOH (final 1 μM, DMSO 0.1% as a cosolvent) in HBSS was loaded into the cells, which were then incubated at 37 °C for 30 min. Without washing, fluorescence images were acquired. In Vivo Imaging. Imaging System. A NightOwl LB 981 molecular light imager (Berthold Technologies, Germany), equipped with a cooled CCD camera (NC 320, Berthold Technologies, Germany), an excitation filter of 630 nm, and an emission filter of 655 nm, was controlled with Berthold WinLight 32 software. Mouse Preparation. Plasmid DNAs (pCMV-SPORT β-gal and pcDNA3.1) were from Invitrogen Corp. (Carlsbad, CA). Male ICR mice (4-week-old, ∼20 g; Japan SLC, Inc.) were given an intravenous (i.v.) injection of saline solution (1.6 mL) containing 10 μg pCMV-SPORT β-gal or pcDNA3.1 via the tail vein over 5 s.20 In Vivo Imaging of Gene Expression. At 24 h after hydrodynamic injection of plasmid DNA, mice were anesthetized by intraperitoneal (i.p.) injection of pentobarbital. A midline abdominal incision was made to expose the liver, and images were acquired (before). The mice were then injected i.v. with 6SqGal (10 μM in 0.1 mL saline), and images were acquired for 30 min. For statistical analysis, fluorescence intensity in the liver was measured with Image J software.

the excited state. We investigated their photochemical properties in the absence or presence of proteins. Most of the dyes showed a red shift and fluorescence increase in the presence of proteins, but the extent of change was strongly dependent on the chemical structure (Table 1). For instance, introduction of propionic acid as a substituent at the N1 position on the indolenine moiety reduced the spectral change, while a benzyl group had little effect (see Cy5, Cy5PH, and Cy5Bn). Moreover, Cy5 with sulfonate groups on the indolenine moiety (Cy5SO3H) exhibited no spectral change (see also Figure S1, Supporting Information). This suggested that negatively charged substituents suppress dye−protein interaction, as we had hypothesized. Squarylium dyes showed a particularly large spectral change; the fluorescence quantum yields were exceedingly small in aqueous buffer in the absence of protein, but increased more than 10-fold in the presence of proteins (Table 1, see also Figure S2). Again, introduction of sulfonate group and propionic acid reduced the extent of spectral change (see SqSO3H and SqSO3H-Me), further supporting the idea that negatively charged substituents suppress the dye−protein interaction. Indeed, the binding parameter nKa [M −1] (Ka, association constant; n, number of independent binding sites on the protein) for Sq with bovine serum albumin (BSA) was 2 orders of magnitude greater than that of SqSO3H (Figures S3 and S4). We also investigated the photochemical properties of the dyes in the presence of positively charged bovine γ-globulin (BGG), in addition to negatively charged BSA. Cyanine and squarylium dyes bearing sulfonate groups exhibited little spectral change, indicating that the effect of negatively charged substituents does not depend on the charge of the protein (Figure S5 and Table S1). Moreover, the absorption and fluorescence spectra in the presence of protein and sodium dodecyl sulfate (SDS, a denaturing agent) were similar to those in the absence of protein (Figure S6), confirming that dye− protein interaction causes the spectral change. With these findings in hand, we focused on the squarylium dye scaffold to develop our strategy of using dye−protein interaction to trigger fluorescence enhancement. As illustrated in Figure 2, the design principle is that the dye is modified with negatively charged functional groups that can be specifically removed by the target enzymes. The released dye should then bind to the adjacent protein with a large enhancement of fluorescence. Development of NIR Fluorescence Probe for Alkaline Phosphatase. We first synthesized three squarylium derivatives (Figure S7), bearing phosphate groups on the indolenine moiety, and measured their spectral change in the presence of protein to further confirm the effect of negatively charged substituents (Figure S8). Since a 5-hydroxyl group on a benzocycle is known to lower the fluorescence quantum yield, presumably by influencing the chromophoric system, we cannot directly compare the fluorescence ratios of the compounds. Therefore, we focused on the changes of absorption and emission wavelength. The extent of these changes was correlated with the amount of negative charge (Table S2), in accordance with our hypothesis that negative charges suppress dye−protein interaction. On the basis of these results, we designed a novel NIR probe for alkaline phosphatase (ALP), which hydrolyzes phosphomonoester to phosphate and a hydroxyl group. ALP is widely used as a marker enzyme in biological studies, including enzyme-linked immunosorbent assays, western blot analysis,21 and gene expression assays,22,23 as well as in clinical diagnosis of

■

RESULTS AND DISCUSSION First, we first synthesized a number of polymethine dyes (Figure 1), including cyanine dyes and squarylium dyes, which we expected to show large fluorescence enhancements in the presence of proteins (we used fetal bovine serum, FBS, which contained 3.6 g/dL protein according to the manufacturer), because they are known to have low fluorescence quantum yields in aqueous media due to fast nonradiative deactivation of

Figure 1. Structures of polymethine dyes. 4406

dx.doi.org/10.1021/ac300061a | Anal. Chem. 2012, 84, 4404−4410

Analytical Chemistry

Article

Table 1. Photochemical Properties of Polymethine Dyes in Phosphate Buffer or FBS

Cy5 Cy5F Cy5MeO Cy5PH Cy5CF3 Cy5SO3H Cy5Bn Sq SqSO3H SqNaphtho SqSO3H-Me 5SqOH 6SqOH

in buffer

in FBS

636 636 656 641 639 646 642 624 629 657 625 637 626

648 642 672 645 647 647 653 638 634 674 633 658 645

ϕFLb

emission maxa (nm)

absorption max (nm) δλ 12 6 16 4 8 1 11 14 5 17 8 21 19

in buffer

in FBS

659 660 685 664 661 669 666 637 644 665 638 658 644

668 666 694 670 666 669 674 650 653 682 648 674 660

δλ 9 6 9 6 5 0 8 13 9 17 10 16 16

in buffer

in FBS

0.11 0.072 0.032 0.16 0.20 0.17 0.11 0.021 0.057 0.013 0.024 0.006 0.015

0.27 0.26 0.087 0.23 0.33 0.20 0.39 0.37 0.26 0.47 0.24 0.14 0.32

ratio 2.5 3.6 2.7 1.4 1.7 1.2 3.5 17.5 4.6 36.0 10.1 23.3 21.3

a Excitation wavelength was 600 nm. bϕFL is the relative fluorescence quantum yield estimated by using ca. 0.1 μM cresyl violet in MeOH (0.54) 19 as a fluorescence standard. δλ is the difference between the absorption or emission maxima in the buffer and in FBS.

Table 2. Photochemical Properties of 6SqPhos and 6SqmonoPhos in Phosphate Buffer or FBS

Figure 2. Design strategy based on change of protein-binding affinity of squarylium dye resulting from specific reaction of the functionalized dye with the target enzymes.

hepatic disease and bone metastasis.24,25 Nevertheless, only a few fluorescence probes for ALP are available, and furthermore these operate in the visible region.26−28 We designed and synthesized 6SqPhos and 6SqmonoPhos by substitution of phosphate for the hydroxyl group(s) of 6SqOH (Figure 3, see

solvent

λabs (nm)

buffer FBS

626 636

buffer FBS

626 640

buffer FBS

626 645

ε ( ×105 M−1 cm−1) 6SqPhos 1.0 0.9 6SqmonoPhos 1.3 0.9 6SqOH 1.1 1.2

λem (nm)

ϕFLa

640 651

0.017 0.26

640 653

0.011 0.25

644 660

0.015 0.32

a ϕFL is the relative fluorescence quantum yield estimated by using ca. 0.1 μM cresyl violet in MeOH (0.54)19 as a fluorescence standard. λabs and λem are the absorption and emission (excitation wavelength was 600 nm) max. ε is the extinction coefficient.

systems), the absorption and emission spectra were red-shifted and the fluorescence was enhanced in an enzyme activitydependent manner, indicating that it should be possible to estimate the quantity of enzyme from a standard curve (Figures 4, S10, and S11). Formation of 6SqOH as a hydrolysis product was confirmed by HPLC analysis (Figure S12). These results demonstrate that 6SqPhos and 6SqmonoPhos are able to detect ALP activity in vitro. Next, we attempted to apply these probes for detection of ALP in western blots. A critical requirement of probes for western blot analysis is that the hydrolysis product has a sufficient affinity for the blot. We first confirmed that 6SqOH has sufficient affinity for the polyvinylidene fluoride (PVDF) membrane after treatment with membrane-blocking agent, while 6SqPhos and 6SqmonoPhos were efficiently washed out with Tris-Tween-buffered saline (Figures S13 and S14). We then examined whether ALP could be visualized on the blot by using 6SqmonoPhos, which showed a more linear response than 6SqPhos, after SDS-polyacrylamide gel electrophoresis and electroblotting. CYP3A2 was adopted as the target protein, and western blot detection was attempted with an ALPconjugated secondary antibody and 6SqmonoPhos. Sharp and CYP3A2-specific red-fluorescent bands were observed, whereas the bands were barely detectable when an HRP-conjugated secondary antibody was used (Figure 5). Moreover, the

Figure 3. Structures of 6SqPhos, 6SqmonoPhos.

also Figure 1 for 6SqOH). The affinity of 6SqPhos for BSA was lower by 1 order of magnitude than that of 6SqOH, which suggests that a protein concentration of a few micromolar would be the minimum required to obtain sufficient fluorescence enhancement, and 6SqPhos showed a lower fluorescence quantum yield in the presence of protein (Table 2 and Figure S9). But, upon addition of ALP to Tris-buffered solutions of 1 μM 6SqPhos or 6SqmonoPhos containing 1% w/ v BSA (corresponding to a protein concentration higher than the minimum required, and closer to that of typical biological 4407

dx.doi.org/10.1021/ac300061a | Anal. Chem. 2012, 84, 4404−4410

Analytical Chemistry

Article

a polymethine dye. We synthesized 6SqGal by substitution of hydroxyl group of 6SqOH with β-galactosyl group (Figure 6).

Figure 6. Structure of 6SqGal.

The binding parameter nKa [M −1] of 6SqGal with BSA was 2 orders of magnitude less than that of 6SqOH, which means that a protein concentration at a few micromolar would be the most appropriate and is the minimum required for sufficient fluorescence enhancement, and 6SqGal exhibited a lower fluorescence quantum yield than 6SqOH in the presence of protein, as shown in Table 3 (see also Figure S18). Upon Table 3. Photochemical Properties of 6SqGal in Phosphate Buffer or FBS

Figure 4. Absorption and fluorescence spectra of 6SqPhos before and after addition of ALP (2.5 units) and time courses of fluorescence intensity of 6SqPhos in the presence of various amounts of ALP. For details, see the procedures detailed in the Experimental Section.

solvent

λabs (nm)

buffer FBS

625 632

buffer FBS

626 645

ε (×105 M−1 cm−1) 6SqGal 2.4 2.0 6SqOH 1.1 1.2

λem (nm)

ϕFLa

640 652

0.017 0.13

644 660

0.015 0.32

a ϕFL is the relative fluorescence quantum yield estimated by using ca. 0.1 μM cresyl violet in MeOH (0.54)19 as a fluorescence standard. λabs and λem are the absorption and emission (excitation wavelength was 600 nm) max. ε is the extinction coefficient.

addition of β-galactosidase to a phosphate-buffered solution of 1 μM 6SqGal containing 1% w/v BSA, in which the protein level is slightly higher than the most appropriate concentration but is closer to that of biological systems, the absorption and emission spectra were red-shifted, and the fluorescence was enhanced in an enzyme activity-dependent manner, indicating that it should be possible to estimate the quantity of enzyme from a standard curve (Figures 7 and S19). Moreover, formation of 6SqOH by hydrolysis of 6SqGal was confirmed by HPLC analysis (Figure S20). These results demonstrate that 6SqGal can detect β-galactosidase activity in vitro. We then applied 6SqGal to HEK293 cells and to HEK293 cells expressing lacZ, the bacterial gene encoding βgalactosidase. Cells (70−80% confluent) were incubated with 6SqGal for 30 min, and fluorescence images were captured. As shown in Figure 8, we obtained bright fluorescence images of lacZ-positive cells, which appeared to be stained diffusely without noteworthy subcellular localization, while lacZ-negative cells showed little fluorescence under the same conditions. On the other hand, both lacZ-positive and lacZ-negative cells were intensely stained with 6SqOH. These results demonstrate that 6SqGal is able to detect β-galactosidase activity in a living cellular system. We also examined the ability of 6SqGal to image βgalactosidase activity in vivo. ICR mice were injected into the tail vein with a β-galactosidase-encoding plasmid targeting the liver.20 After 24 h, the mice were anesthetized, and a midline abdominal incision was made to expose the liver. Then, the mice were intravenously injected with 6SqGal, and fluorescence

Figure 5. Western blot detection of CYP3A2 using HRP-conjugated secondary antibody (left) or ALP-conjugated secondary antibody (right) with 6SqmonoPhos. The seven lanes contain protein molecular weight standard (lane 1) and decreasing amounts of CYP3A2 (0.2 pmol to 0.01 pmol, lane 2 to lane 7). Data from two representative gels are shown (6 gels were run in 3 independent experiments). WL, white light; FL, fluorescence; CL, chemiluminescence.

fluorescence intensity of each band was well correlated with the amount of protein (Figure S15). We also compared the usefulness of 6SqmonoPhos with that of a commercially available ALP-detection reagent, DDAO-Phos. 6SqmonoPhos exhibited an approximately 2-fold higher signal than DDAOPhos, presumably reflecting the affinity for proteins blocking the membrane, while the background was in the same range (Figure S16, see also Figure S17). These results demonstrate that 6SqmonoPhos is an effective detection reagent for ALPconjugated secondary antibodies in Western blot analysis. Development of an NIR Fluorescence Probe for βGalactosidase. We expected that hydrophilic groups, such as sugars, would also suppress dye−protein interaction and so could be used to modulate the protein binding affinity of NIR probes according to our design strategy. Among glycosidases, βgalactosidase is the most widely used marker enzyme in enzyme-linked immunosorbent assays,29,30 in situ hybridizations,31,32 and gene expression assays,33,34 and many fluorescence probes for β-galactosidase have been developed.35−39 However, so far there is no such probe based on 4408

dx.doi.org/10.1021/ac300061a | Anal. Chem. 2012, 84, 4404−4410

Analytical Chemistry

Article

(see also Figure S22). These results demonstrate that 6SqGal can be used for in vivo imaging of β-galactosidase activity.

■

CONCLUSION We have developed a novel design strategy for near-infrared (NIR) fluorescence probes, based on modification of a polymethine dye with cleavable functional groups that reduce the dye’s protein-binding affinity, with a concomitant decrease of fluorescence. When the functional groups are removed by specific interaction with the target enzymes, the dye’s proteinbinding affinity is restored, protein binding occurs, and the dye’s fluorescence is enhanced. As proof of concept, we synthesized the first squarylium dye-based fluorescent sensor for ALP, and confirmed its suitability for detecting ALP-labeled secondary antibodies in western blotting analysis. We also synthesized a probe for β-galactosidase (6SqGal), which is widely used as a reporter of gene expression, and confirmed that the probe can visualize β-galactosidase activity expressed in HEK293 cells in vitro and in mouse liver in vivo. On the basis of these results, we anticipate that our design strategy will be a powerful tool for designing a wide range of functional NIR fluorescence probes.

Figure 7. Absorption and fluorescence spectra of 6SqGal before and after addition of β-galactosidase (12 units) and time courses of fluorescence intensity observed with 6SqGal in the presence of various amounts of β-galactosidase. For details, see the procedures detailed in the Experimental Section.

■

ASSOCIATED CONTENT

S Supporting Information *

Experimental procedures and supporting figures. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone: +81-3-5841-4850. Fax: +81-3-5841-4855. Notes

The authors declare no competing financial interest.

■

Figure 8. Fluorescence and differential interference contrast (DIC) images of lacZ-positive or lacZ-negative HEK293 cells loaded with 6SqGal or 6SqOH. Representative images are shown (n = 3). DIC, differential interference contrast; FL, fluorescence.

ACKNOWLEDGMENTS This work was in part supported by a Grant-in-Aid for JSPS Fellows (to D.O.), by a Grant-in-Aid for Specially Promoted Research (Grant 22000006 to T.N.), and a Grant-in-Aid for Scientific Research (Grant No 23651231 to T.T.) by the Ministry of Education, Culture, Sports, Science and Technology of Japan, and by funding from New Energy and Industrial Technology Development Organization (NEDO) of Japan (to T.T.). T.T. was also supported by The Cosmetology Research Foundation. K.H. was supported by Inoue Foundation for Science, Konica Minolta Science and Technology Foundation, The Asahi Glass Foundation, and Takeda Science Foundation.

images were acquired sequentially for 30 min. As shown in Figure 9, the fluorescence intensity in the liver was greatly increased in the β-galactosidase-expressing animals (see also Figure S21). The fluorescence enhancement was hardly observed in mice inoculated with a negative control plasmid

■

REFERENCES

(1) Weissleder, R. Nat. Rev. Cancer 2002, 2, 11−18. (2) Rudin, M.; Weissleder, R. Nat. Rev. Drug Discov. 2003, 2, 123− 131. (3) Ueno, T.; Nagano, T. Nat. Methods 2011, 8, 642−645. (4) Weissleder, R. Nat. Biotechnol. 2001, 19, 316−317. (5) Weissleder, R.; Ntziachristos, V. Nat. Med. 2003, 9, 123−128. (6) Frangioni, J. V. Curr. Opin. Chem. Biol. 2003, 7, 626−634. (7) Ozmen, B.; Akkaya, E. U. Tetrahedron Lett. 2000, 41, 9185−9188. (8) Sasaki, E.; Kojima, H.; Nishimatsu, H.; Urano, Y.; Kikuchi, K.; Hirata, Y.; Nagano, T. J. Am. Chem. Soc. 2005, 127, 3684−3685. (9) Kiyose, K.; Kojima, H.; Urano, Y.; Nagano, T. J. Am. Chem. Soc. 2006, 128, 6548−6549. (10) Myochin, T.; Kiyose, K.; Hanaoka, K.; Kojima, H.; Terai, T.; Nagano, T. J. Am. Chem. Soc. 2011, 133, 3401−3409.

Figure 9. Comparison of mice inoculated with pCMV-SPORT β-gal (top) or pcDNA3.1 (bottom). Representative merged images of the white light and fluorescence images are shown (n = 3). 4409

dx.doi.org/10.1021/ac300061a | Anal. Chem. 2012, 84, 4404−4410

Analytical Chemistry

Article

(11) Blum, G.; von Degenfeld, G.; Merchant, M. J.; Blau, H. M.; Bogyo, M. Nat. Chem. Biol. 2007, 3, 668−677. (12) Weissleder, R.; Tung, C. H.; Mahmood, U.; Bogdanov, A., Jr. Nat. Biotechnol. 1999, 17, 375−378. (13) Xing, B.; Khanamiryan, A.; Rao, J. J. Am. Chem. Soc. 2005, 127, 4158−4159. (14) Kiyose, K.; Hanaoka, K.; Oushiki, D.; Nakamura, T.; Kajimura, M.; Suematsu, M.; Nishimatsu, H.; Yamane, T.; Terai, T.; Hirata, Y.; Nagano, T. J. Am. Chem. Soc. 2010, 132, 15846−15848. (15) Oushiki, D.; Kojima, H.; Terai, T.; Arita, M.; Hanaoka, K.; Urano, Y.; Nagano, T. J. Am. Chem. Soc. 2010, 132, 2795−2801. (16) Awasthi, K.; Nishimura, G. Photochem. Photobiol. Sci. 2011, 10, 461−463. (17) Losytskyy, M. Y.; Volkova, K. D.; Kovalska, V. B.; Makovenko, I. E.; Slominskii, Y. L.; Tolmachev, O. I.; Yarmoluk, S. M. J. Fluoresc. 2005, 15, 849−857. (18) Tatikolov, A. S.; Costa, S. M. Biophys. Chem. 2004, 107, 33−49. (19) Magde, D.; Brannon, J. H.; Cramers, T. L.; Olmsted, J., III. J. Phys. Chem. 1979, 83, 696−699. (20) Liu, F.; Song, Y.; Liu, D. Gene Ther. 1999, 6, 1258−1266. (21) Reen, D. J. Methods Mol. Biol. 1994, 32, 461−466. (22) Cullen, B. R.; Malim, M. H. Methods Enzymol. 1992, 216, 362− 368. (23) Weber, W.; Rimann, M.; Spielmann, M.; Keller, B.; Daoud-El Baba, M.; Aubel, D.; Weber, C. C.; Fussenegger, M. Nat. Biotechnol. 2004, 22, 1440−1444. (24) Bodlaj, G.; Hubmann, R.; Saleh, K.; Stojakovic, T.; Biesenbach, G.; Berg, J. World J. Gastroenterol. 2010, 16, 2407−2410. (25) Kataoka, A.; Yuasa, T.; Kageyama, S.; Tsuchiya, N.; Habuchi, T.; Iwaki, H.; Narita, M.; Okada, Y.; Yoshiki, T. Clin. Oncol. (R. Coll. Radiol.) 2006, 18, 480−484. (26) Kamiya, M.; Urano, Y.; Ebata, N.; Yamamoto, M.; Kosuge, J.; Nagano, T. Angew. Chem., Int. Ed.. 2005, 44, 5439−5441. (27) Kawaguchi, M.; Hanaoka, K.; Komatsu, T.; Terai, T.; Nagano, T. Bioorg. Med. Chem. Lett. 2011, 21, 5088−5091. (28) Kim, T. I.; Kim, H.; Choi, Y.; Kim, Y. Chem. Commun. 2011, 47, 9825−9827. (29) Konijn, A. M.; Levy, R.; Link, G.; Hershko, C. J. Immunol. Methods 1982, 54, 297−307. (30) Baruch, D.; Glickstein, H.; Cabantchik, Z. I. Exp. Parasitol. 1991, 73, 440−450. (31) Fields, S.; Sternglanz, R. Trends Genet. 1994, 10, 286−292. (32) Vidal, M.; Brachmann, R. K.; Fattaey, A.; Harlow, E.; Boeke, J. D. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 10315−10320. (33) Lin, S.; Yang, S.; Hopkins, N. Dev. Biol. 1994, 161, 77−83. (34) Berger, C. N.; Tan, S. S.; Sturm, K. S. Cytometry 1994, 17, 216− 223. (35) Rakhmanova, V. A.; MacDonald, R. C. Anal. Biochem. 1998, 257, 234−237. (36) Corey, P. F.; Trimmer, R. W.; Biddlecom, W. G. Angew. Chem., Int. Ed. Engl. 1991, 30, 1646−1648. (37) Tung, C. H.; Zeng, Q.; Shah, K.; Kim, D. E.; Schellingerhout, D.; Weissleder, R. Cancer Res. 2004, 64, 1579−1583. (38) Kamiya, M.; Kobayashi, H.; Hama, Y.; Koyama, Y.; Bernardo, M.; Nagano, T.; Choyke, P. L.; Urano, Y. J. Am. Chem. Soc. 2007, 129, 3918−3929. (39) Kamiya, M.; Asanuma, D.; Kuranaga, E.; Takeishi, A.; Sakabe, M.; Miura, M.; Nagano, T.; Urano, Y. J. Am. Chem. Soc. 2011, 133, 12960−12963.

4410

dx.doi.org/10.1021/ac300061a | Anal. Chem. 2012, 84, 4404−4410